Decision Tree Based Customer Analysis Method for Energy Planning in Smart Cities

Abstract

In this study, a method is proposed to estimate energy consumption and to plan maintenance works on energy lines according to energy consumption. In our study, energy consumption can be estimated at any time of the day by analysing parameters such as temperature, pressure, and wind using decision tree methods. For this aim, "Fine Tree", "Medium Tree"and "Coarse Tree"which are different implementations of decision trees are used. The success is 76.77% with the Fine Tree method, 76.42% with the Medium Tree method, and 73.92% with the Coarse Tree method. As a result, Energy consumption is estimated by using environmental conditions, time, day, month, and season information within the scope of the smart city. Thus, energy distribution companies can make energy planning by estimating the energy consumption of their customers. In addition, performing maintenance and repair operations on energy distribution lines in a convenient time period according to these results will increase customer satisfaction. © 2020 IEEE.